HAL Id: hal-00634887
https://hal.archives-ouvertes.fr/hal-00634887
Submitted on 24 Nov 2016
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Tertiary sequence of deformation in a
thin-skinned/thick-skinned collision belt: The Zagros
Folded Belt (Fars, Iran)
F. Mouthereau, J. Tensi, N. Bellahsen, O. Lacombe, T. de Boisgrollier, S.
Kargar
To cite this version:
F. Mouthereau, J. Tensi, N. Bellahsen, O. Lacombe, T. de Boisgrollier, et al.. Tertiary sequence of deformation in a thin-skinned/thick-skinned collision belt: The Zagros Folded Belt (Fars, Iran). Tectonics, American Geophysical Union (AGU), 2007, 26 (5), �10.1029/2007TC002098�. �hal-00634887�
Tertiary sequence of deformation in a thin-skinned/thick-skinned
collision belt: The Zagros Folded Belt (Fars, Iran)
F. Mouthereau,1J. Tensi,1N. Bellahsen,1O. Lacombe,1T. De Boisgrollier,1and S. Kargar2
Received 3 January 2007; revised 14 June 2007; accepted 5 July 2007; published 27 September 2007.
[1] We describe how thin-skinned/thick-skinned
deformation in the Zagros Folded Belt interacted in time and space. Homogeneous fold wavelengths (15.8 ± 5.3 km), tectono-sedimentary evidence for simultaneous fold growth in the past 5.5 ± 2.5 Ma, drainage network organization, and homogeneous peak differential stresses (40 ± 15 MPa) together point to buckling as the dominant process responsible for cover folding. Basin analysis reveals that basement inversion occurred 20 Ma ago as the Arabia/Eurasian plate convergence reduced and accumulation of Neogene siliciclastics in foreland basin started. By 10 Ma, ongoing contraction occurred by underplating of Arabian crustal units beneath the Iranian plate. This process represents 75% of the total shortening. It is not before 5 Ma that the Zagros foreland was incorporated into the southward propagating basement thrust wedge. Folds rejuvenated by 3– 2 Ma because of uplift driven by basement shortening and erosion. Since then, folds grew at 0.3—0.6 mm/yr and forced the rivers to flow axially. A total shortening of 65– 78 km (16 – 19%) is estimated across the Zagros. This corresponds to shortening rates of 6.5 – 8 km/Ma consistent with current geodetic surveys. We point out that although thin-skinned deformation in the sedimentary cover may be important, basement-involved shortening should not be neglected as it requires far less shortening. Moreover, for such foreland folded belts involving basement shortening, underplating may be an efficient process accommodating a significant part of the plate convergence. Citation: Mouthereau, F., J. Tensi, N. Bellahsen, O. Lacombe, T. de Boisgrollier, and S. Kargar (2007), Tertiary sequence of deformation in a thin-skinned/thick-skinned collision belt: The Zagros Folded Belt (Fars, Iran), Tectonics, 26, TC5006, doi:10.1029/2007TC002098.
1. Introduction
[2] In an attempt at illustrating the structure and kine-matics of foreland fold-and-thrust belts, the most popular
model, thin-skinned in style, considers it as the product of continuous subduction of the lower plate basement beneath the upper plate. There are several limitations for using such a thin-skinned deformation model for collisional settings as it assumes that a significant part of the total plate conver-gence might be accommodated at the front, by rocks accretion above a sole thrust at shallow crustal levels and hence neglects intraplate basement thickening, giving rise to unrealistically high cover shortening. The alternative thick-skinned interpretation is supported by our current knowl-edge on the rheology of the continental crust: it is buoyant and rheologically stratified, including a ductile lower crust, which allows for decoupling the upper crust from the mantle. Moreover, the development of foreland fold-and-thrust belts over rifted continental margin likely promotes reverse-reactivation/inversion of preorogenic basement nor-mal faults. Such basement-involved deformation simply highlights plate coupling in collisional settings and requires far less shortening than equivalent thin-skinned models do. [3] Numerous examples exist of external fold-and-thrust belts in which basement is known to be involved in the compressional deformation such as the Andes [Zapata and Allmendinger, 1996; Cristallini and Ramos, 2000], the northern Apennines [Coward et al., 1999; Butler et al., 2004], Taiwan [Lee et al., 2002; Mouthereau and Lacombe, 2006], the Alps [Meyer et al., 1994; Lacombe and Mouthereau, 2002; Rostein and Schaming, 2004], the Zagros [Sherkati and Letouzey, 2004; Molinaro et al., 2005]. In spite of increasing evidence supporting the coexistence of both thin-skinned and thick-skinned inter-pretation at belt fronts, the interplay in time and space of both deformations has received, to date, little attention. However, our understanding of the overall kinematic evo-lution of mountain belts is dependent on our understanding of how and when contraction of the basement is transferred up-section to the cover sequence folding.
[4] The Zagros folded belt offers a unique opportunity to examine such interactions. The presence of a thick and weak salt-bearing formation at the base of the sedimentary cover and fold geometry has led for years to interpret the structural style in the Zagros Folded Belt in thin-skinned style [Davis and Engelder, 1985]. Since crystalline base-ment is generally not exposed in forelands, the recognition of underlying basement-involved deformation is usually indirect [e.g., Lacombe and Mouthereau, 2002]. In the northwest of the Zagros Folded Belt, basement-involved shortening was required for balancing the differential uplift between adjacent synclines [Blanc et al., 2003; Sherkati and Letouzey, 2004]. This is probably the present-day crustal deformation patterns including seismotectonic studies that
Here for Full Article
1Laboratoire Tectonique, Universite´ Pierre et Marie Curie, UMR 7072,
Paris, France.
2Geological Survey of Iran, Tehran, Iran.
Copyright 2007 by the American Geophysical Union. 0278-7407/07/2007TC002098$12.00
provide the more reliable evidence for thick-skinned defor-mation [Jackson, 1980; Ni and Barazangi, 1986; Berberian, 1995]. It is no doubt that both structural styles currently coexist in the Zagros Folded Belt, but to first order this is the growth of a thrust wedge including basement-involved deformation that is mechanically required to maintain the current topography [Mouthereau et al., 2006].
[5] Our goal in this paper is to examine how thin-skinned deformation and thick-skinned deformation in the Zagros Folded Belt have interacted in time and space since the Middle Miocene. To this aim we provide a new crustal-scale section across the Central Fars area of the Zagros Folded Belt. We use isopach maps to assess the large-scale tecton-ically induced movements, which have controlled the sed-imentation in the Zagros foreland basin. Unconformities reported from fieldwork within the synorogenic Neogene strata of the synclinal valleys offer direct constraints re-garding the inception and the growth of cover folding. In order to decipher the last Quaternary increment of shorten-ing, rates of folding and sequence of deformation between thick-skinned and thin-skinned mode, we examine folds geomorphology and associated drainage networks. In the absence of currently reliable stratigraphic ages of the synorogenic continental deposits, age constraints are brought by the extrapolation of synorogenic accumulation rates in the Zagros Folded Belt to the observed thickness of strata with unconstrained ages. These data are used to give preliminary indicative estimates for fold uplift rates aver-aged over the past 3 – 2 Ma, which have been poorly constrained so far. With the aim at offering a viable mechanical explanation of the sequence of deformation, we discuss the possibility that cover folding may result primarily from the growth of buckling instabilities rather than fault-related folding. On the basis of the restored regional crustal-scale section and our results on the se-quence of deformation we draw a forward scenario of the Tertiary evolution of the Zagros Folded Belt from basin inversion, inception of thin-skinned fold growth to the current basement-involved thrusting. All these results finally place constraints on the topographic evolution of the Zagros orogeny and potential Neogene climatic forcing in the framework of the Arabia/Eurasia plate convergence.
2. Central Fars: Geological Background
[6] The Zagros Mountains form a broad orogenic domain in Iran, approximately 2000 km long, in front of the Turkish-Iranian plateau (Figure 1). The mountain belt results from the ongoing accommodation of the conver-gence between the rifted continental margin of the Arabian plate and the Iranian continental blocks following the closure of the Neo-Tethys ocean during the Tertiary [Stocklin, 1968; Koop and Stoneley, 1982]. The present-day convergence between Arabia and Eurasia is 2 – 3 cm/yr [Vernant et al., 2004] in the N – S direction and is assumed to be unchanged since at least 10 Ma [McQuarrie et al., 2003]. The collision suture zone is outlined, along the Main Zagros Thrust (MZT), by a complex of ophiolitic rocks, deep-water radiolarites and eruptive rocks interpreted asremnants of the obducted Neo-Tethyan ocean or one of its derivative (e.g., back-arc or fore-arc crust) [Stoneley, 1990; Ziegler, 2001]. The tectono-metamorphic belt of Sanandaj-Sirjan North of the MZT represents the former active margin of the Iranian microplate (Figure 1). To the South, the Imbricate Zone (or High Zagros) and the Zagros Folded Belt (ZFB) form a large external folded domain within the rifted Arabian continental margin.
[7] In this paper we focus on the Central Fars province of the Zagros Folded Belt (Figures 1 and 2). It is a 200-km-wide arcuate folded belt, south of the High Zagros Fault, characterized by periodic, symmetrical and open concentric folds with axial lengths often greater than 100 km. The ZFB results from the folding of a thick pile of sedimentary rocks up to 12 km [Stocklin, 1968; Falcon, 1974; Colman-Sadd, 1978] including Paleozoic, Mesozoic strata and Cenozoic synorogenic deposits (Figures 2 and 3). The exceptional geomorphic expression of folds, the so-called ‘‘whaleback’’ shape of folds, is related to the presence, within the upper part of the folded pile, of resistant carbonates belonging to the Asmari Formation, Oligo-Miocene in age. Noticeably, the topographic elevation of the crest of anticlines above local base levels is remarkably uniform.
3. Crustal-Scale Cross Section of the Zagros
(Central Fars)
3.1. Subsurface Constraints on the Crustal Structure of the Zagros Folded Belt and Internal Domains
[8] In order to examine the sequence of deformation and to estimate finite shortening in both the cover and the basement we have constructed a crustal-scale balanced section across the Zagros Folded Belt, the High Zagros and the Sanandaj-Sirjan Zone (Figures 2 and 4). Surface geology was constrained by field studies, Landsat images and geological maps from the National Iranian Oil Compa-ny (1:1,000,000 and 1:250,000 scale maps). Subsurface constraints in the ZFB were provided by the thickness distribution of Paleozoic, Mesozoic and Cenozoic strata from available isopachs and basement isobaths derived from aeromagnetic surveys [Motiei, 1993].
[9] Our cross section extends northward into the Sanan-daj-Sirjan area. However, it was beyond our objectives to study the structure of the SSZ, so we treat this folded domain as a coherent assemblage of tectono-metamorphic units belonging to the upper Iranian plate. The suture zone, which is represented by the ophiolites of Neyriz, is exposed just 50 km east of the studied section (Figure 2). This ophiolitic complex represents a klippen of allochthonous radiolarites, peridotites and exotic masses of Mesozoic carbonate overthrusted onto the continental margin in the Late Cretaceous (Figure 3). The oceanic basin has now disappeared but we expect that ophiolitic units are present below the SSZ units. This is suggested by the geometry of the Moho provided by recent geophysical studies and receiver functions revealing that the Arabian crust and possibly the oceanic crust have been underplated below the MZT [Hatzfeld et al., 2003; Paul et al., 2006]. The
depth of the mechanical lithosphere is set to 100 km in slightly stretched areas of the Arabian continental margin, which corresponds to the depth of the 700° – 800°C iso-therm [Burov and Diament, 1995].
3.2. Superimposed Thin-Skinned and Thick-Skinned Deformation
[10] Figure 4 clearly shows that deep-seated thick-skinned deformation is overlain by thin-thick-skinned
deforma-tion mode at a shallower crustal level. The preservadeforma-tion of layer area required in balancing is solved by filling the core of anticlines by the 2-km-thick ductile Hormuz Formation and/or by basement culminations outlining the thick-skinned deformation style.
[11] Numerous features of the ZFB including low topographic slopes [Talbot and Alavi, 1996], symmetrical anticlines and the lack of clear fold vergence have been referred to as the expression of a brittle thin-skinned wedge Figure 1. Zagros Folded Belt in the framework of the Arabia/Eurasia plate convergence. Black arrows
indicate the present-day convergence of the Arabian plate with respect to stable Eurasia [De Mets et al., 1994]. It predicts a present-day convergence of3 ± 0.5 cm/yr oriented N – S on average at the front of the Zagros Mountains. The studied area of the Central Fars province is indicated by a black quadrangle. The suture zone is outlined by remnants of ophiolitic series and the Main Zagros Thrust (MZT). The inset in the lower left shows seismicity in the Zagros for earthquakes with focal depths shallower than 35 km and magnitudes 2.4 < mb< 7.4 (data issued from ISC and CMT catalogs, years 1965 – 2003). Capital letters L, D, and F for Lorestan, Dezful, and Fars areas, respectively.
Figure 2. Geological map of the Fars based on a compilation of geological 1:250,000 and 1:100,000 scale maps from the National Iranian Oil Company [1977]. The Imbricate Zone (High Zagros) north of the High Zagros Fault is characterized by the lack of Oligo-Miocene deposits which contrast with the Zagros Folded Belt to the south. The anticlines and the major transverse strike-slip fault zone mentioned in the present are depicted. The thick black solid line corresponds to the location of the crustal-scale section of Figure 4. Two compressive focal mechanisms after [Talebian and Jackson, 2004] located to the north of the Surmeh Fault have been used to constrain the geometry at depth of the Surmeh basement thrust.
Figure 2
overlying a ductile de´collement [Davis and Engelder, 1985]. The structural styles of the ZFB in the Central Fars, however, differs from other salt-based wedges like the Jura folded belt [Sommaruga, 1999] or the Salt Ranges in Pakistan [Jaume´ and Lillie, 1988] in several points. First, there is a lack of unambiguous evidence for thrust-related folding (Figures 2 and 4). Second, the ratio of the maximum
folds amplitude (500 – 1000 m [see Mouthereau et al., 2006, Figure 8]) to the fold wavelength (15.8 ± 5.3 km; see Figure 5) is remarkably constant throughout the belt indi-cating insignificant deformation gradient. The lack of increasing shortening toward the rear of the ZFB is in agreement with the nearly homogeneous Mio-Pliocene peak differential stresses values (40 ± 15 MPa), which were determined from analyses of calcite twinning across the Simply Folded Belt in synfolding veins and host rocks [Lacombe et al., 2007]. This suggests that cover folding might be primarily linked to pure-shear deformation and buckling of the sedimentary rocks rather then brittle tectonic wedging as suggested by recent studies [Schmalholz et al., 2002; Mouthereau et al., 2007]. Furthermore, the space problem related to the projection to depth of concentric folds may be solved assuming that part of the cover shortening is taken up by plastic and viscous internal deformation such as pressure solution or calcite twinning [Lacombe et al., 2007; Mouthereau et al., 2007]. The quasiperfect sinusoidal shape of cover folds appears to be essentially superimposed onto the larger-scale regional topographic slope originated in the basement. As a result, instead of the common feature of a weak basal de´collement dipping toward the hinterland, the Hormuz de´collement regionally dips toward the foreland at the top of the basement wedge.
[12] The way basement deforms is however significantly different since faulting seems to be an important, but not unique, mechanism accommodating contraction in the base-ment as shown by the seismicity [Berberian, 1995; Talebian and Jackson, 2004; Tatar et al., 2004]. The observation of the lack of basement seismogenic strain rate with respect to cover geodetic strain rate however provides evidence for longer-term ductile and aseismic deformation in the crust [Jackson and McKenzie, 1988; Masson et al., 2004]. The location and the dip of major basement-involved thrusts in Figure 4 are provided by the study of historical earthquakes and current focal mechanisms reported in this area [Ambraseys and Melville, 1982; Berberian, 1995]. For instance, the dip of the Surmeh thrust was constrained by focal mechanisms of two earthquakes with magnitudes Mw higher than 4.6 [Talebian and Jackson, 2004]. Both of them occurred at depth shallower than 15 km and are consistent with reverse motion along the moderately dipping Surmeh thrust. The depth at which decoupling between the upper brittle crust and the underlying ductile lower crust likely occurs can be used as an additional constraint to infer the position of the low-angle ductile reverse shear zone; the depth of decoupling lies between 33 and 39 km given a Moho depth at 45 km [Mouthereau et al., 2006]. Beneath the Imbricate Zone, a third basement-involved thrust is proposed in order to explain the exposure of Mesozoic rocks in the absence of cover thrusting.
[13] The displacement across active basement faults has been indirectly estimated by vertical offset of local base levels. The total vertical movement within the basement, which is accommodated by reverse faults, is consistent with the cumulated regional topographic elevation of 2500 m (equivalent to a topographic slope of 0.5° on average). By Figure 3. Simplified chronostratigraphic chart and
lithol-ogies encountered in the Central Fars after Motiei [1993]. The basal de´collement (largest black arrow) of the folded cover lies within the Eo-Cambrian salts of the Hormuz formation. Other smaller black arrows refer to second-order and shallower layers that may act as potential detachment levels in the Zagros Folded Belt. Major unconformities are depicted to emphasize the main periods of tectonic activity experienced by the Arabian margin since rifting to collision and uplift of the Zagros.
Figur
e
4
means of horizontal contraction, and assumption on faults dip, the observed uplift corresponds to shortening of about 15 km. Excess of outward propagation along basement thrusts with respect to the overlying cover allows for differential plastic deformation possibly accommodated by triangle zones within the Hormuz Salt (Figure 4).
[14] A noticeable difference from previous cross sections in the Fars [McQuarrie, 2004] is the lack of thrusts associated with cover folding. If microscale and mesoscale faulting does occur in the sedimentary cover [Lacombe et al., 2006], at larger scale, evidence for fold-related thrusting is generally lacking in the field and not supported by the depth of earthquakes mostly located in the basement [e.g., Talebian and Jackson, 2004]. The sedimentary cover is however cut by major thrusts exactly where seismogenic fault zones are reported in the basement such as above the Mountain Front Fault, the Surmeh-Karebass and
Sabz-Pushan transverse fault zones. Another evidence is that spacing for fold-related thrusts imposed by the dominant folds wavelength (Figure 5) is too small for fitting internal friction angle in classical brittle thin-skinned wedge [Mouthereau et al., 2006]. We do not exclude, however, as shortening increases and fold limbs rotate that thrusting occurred to accommodate plastic strain in folds hinges. 3.3. Role of Intermediate Cover De´collement Levels
[15] Another difference in regard to the Dezful-Izeh area [Sherkati and Letouzey, 2004] and the Bandar-Abbas area [Molinaro et al., 2005] is that intermediate cover de´colle-ment levels are lacking in our section. Though the well-known Neogene evaporites-bearing layers of the Gashsaran Formation are present in the southern Fars, no layer-parallel shear has been reported, so it is probably not acting as an intermediate detachment in the studied area. Moreover, the Figure 5. Distribution of fold wavelengths measured for 149 anticlines (modified after Mouthereau et
al. [2007]). Fold widths have been estimated as the distance between two adjacent synclines independently from the stratigraphy of the folded strata. The diagram shows the calculated statistical Gaussian distribution. It reveals that Zagros folds have a dominant wavelength of 15.8 ± 5.3 km.
Figure 4. Crustal-scale section of the Zagros Folded Belt. Subsurface constraints in the ZFB were provided by the thickness distribution of Paleozoic, Mesozoic, and Cenozoic strata from available isopachs and basement isobaths derived from aeromagnetic surveys [Motiei, 1993]. The Sanandaj-Sirjan Zone (SSZ) is shown as a coherent assemblage of tectono-metamorphic units belonging to the upper Iranian plate. The currently inactive suture zone is represented by the ophiolites of Neyriz. Tectonic underplating beneath the SSZ area is suggested by the geometry of the Moho provided by recent geophysical studies and receiver functions [Hatzfeld et al., 2003; Paul et al., 2006]. The depth of the mechanical lithosphere is set to 100 km in the unbended and slightly stretched areas of the Arabian continental margin, i.e., at the Mountain Front, which corresponds to the depth of the 700° – 800°C isotherm [Burov and Diament, 1995]. Exaggerated topography and differential stresses that derived from calcite twinning analyses are shown above [Lacombe et al., 2007].
existence of an intermediate potential detachment, for example, the evaporitic horizon of the Triassic Dashtak Formation (Figure 3), would lead to uniform reduction of the wavelengths of folding, which is not observed either in the field or based on subsurface data [Sherkati et al., 2005; Oveisi et al., 2007]. Intermediate detachment layers appear of second-order importance to explain the observed domi-nant wavelength, the style of folding being primarily controlled by the thickness of a single competent unit and the decoupling above a single basal de´collement layer lying in the Hormuz Formation (Figure 5).
3.4. Restoration and Estimates of Crustal Shortening [16] Across both the Zagros Folded Belt and the Imbri-cate Zone, we estimate only15 km (5%) of horizontal shortening taken up in both the cover and the basement. In our interpretation, restoration of cover thrusts show clear structural relationships with ancient NW – SE-trending extensional faults inherited from the Tethyan rifting.
[17] The geometry of the Moho deduced from the tomography [Paul et al., 2006] shows that the Arabian
margin was tectonically thickened and underplated beneath the Main Zagros Thrust. To account for this observation, thrust sheets including basement and cover units are likely to be root beneath the MZT. These units presumably correspond to former half-grabens accommodating the thin-ning of the passive continental margin, secondarily inverted when orogenic contraction began. We estimate 50 km of shortening related to underplating of Arabian crust beneath the Iranian plate. In this scenario, the current position of the ophiolitic units is found 50 km north of the MZT beneath the SSZ. The thickness and nature of the sedimentary cover deposited above the continental rise and oceanic crust, together with syn-obduction strata deposited in the footwall of the overthrusted oceanic crust is however unknown. Indeed, they were probably removed by erosion and/or underplated beneath the ophiolitic units.
[18] In order to account for penetrative deformation we consider that possibly up to 30% of plastic-viscous internal deformation may occur in agreement with some estimates obtained elsewhere [Sans et al., 2003]. The section thus accounts for a total shortening of 65 km (16%) or 78 km (19%) considering 30% of additional shortening. These Figure 6. Stratigraphy of the Oligocene and Neogene strata contemporaneous with the Zagros orogeny
(modified after James and Wynd [1965]). Black dashed line outlines the southward onlap progressive in time of the shallowing-upward synorogenic deposition (Razak-Gashsaran Formation) onto the carbonates of the Asmari Formation in the context of flexural basin development. Eustatic sea level curve is after Haq et al. [1987].
values are close to another estimate of 17% proposed in the same region [McQuarrie, 2004].
4. Foreland Neogene Sequences: Evidence
for Early Miocene Margin Inversion
[19] To evaluate the possibility for margin inversion during the early stages of the collision, we examine the relation in space between thickening/thinning of the syn-orogenic siliciclastic deposits and the current position of large-scale basement faulting within the basin. Sedimento-logical data, isopach maps based on a compilation per-formed by Motiei [1993] and stratigraphic constraints provided by James and Wynd [1965] and Homke et al. [2004] are combined along with information from fieldwork. [20] Foreland basins form by the flexure of the litho-sphere in front of advancing thrust-fold belts or under a combination of orogenic and subcrustal loads [e.g., Watts, 2001]. The inception of foreland deposition is often found in association with an unconformity that is attributed to the rapid subsidence and drowning of precollisional margin sequences. As the collision migrates, the synorogenic silici-clastic foreland sequences are characterized by onlaps of the margin sequences and a deepening-upward sequence above the foreland unconformity.
[21] In the Zagros Basin, foreland sequences of Miocene ages are represented by a thick regressive siliciclastic sequence (up to 3000 m), namely, the Fars Group lying above the well-developed carbonate platform of the Asmari Formation (Figures 2 and 6). Figure 6 shows that the initiation of foreland siliclastics deposition is delayed to-ward the foreland: the onset of clastics deposition is Chattian (28 Ma) in the northern Fars whereas it is Burdigalian (20 – 16 Ma) near the Persian Gulf. The prograding sequences of the Fars Group likely correspond to the typical overfilled stage often reported in foreland basins [Sinclair, 1997]. The inception of foreland transgres-sive sequences above margin strata is however not clearly identified from the sedimentary patterns (Figure 7) though
the carbonates of the Asmari Formation may represent part of this initial transgressive sequence. We challenge that most of the fining-upward foreland sequences have been deposited northward then eroded and/or buried below the MZT.
[22] Isopachs maps of Figure 8 show an inhomogeneous pattern of deposition. The thickness of the Fars Group is maximum to the north owing to plate flexure. Second-order NW – SE-trending depocenters reveal a complex pattern of subsidence in relation with tectonic inheritance and base-ment block faulting in the Arabian margin. The deposition of the Fars Group in the foreland domain coincides with a significant increase in accumulation rates up to 182 m/Ma over the past 12 Ma (Figure 9).
4.1. Early Miocene: Gashsaran/Razak Formations [23] The Lower Miocene is represented by the Gashsaran and the Razak formations in the southern and the northern parts of the studied area, respectively. They form a prograding sequence overlying the Asmari limestones (Figures 6 and 7). Though fieldwork did not bring evidence for unconformities at the base of these series, the Gashsaran Formation and the lower Mishan Formation are locally missing above the Asmari carbonates, near the coast of the Persian Gulf, on the southern limb of the Asaluyeh anticline [Mouthereau et al., 2006].
[24] In the northern Zagros, the Razak Formation consists of 900-m-thick red shales with nodular anhydrites, typical of intertidal to supratidal sabkhas environments [Tucker, 2005]. The local abundance of red siltstones and sandstones beds with megaripples shows that the sedimentation is controlled by clastic sediment supplies (Figure 10a). South-ward, the thickness of the Gashsaran/Razak formations reduces to 600 m above a paleo-bathymetric high. Near Firuz Abad (labeled F in Figure 8), the thickness of the Gashsaran/Razak Formation increases to 900 m. These deposits are characterized by deeper intertidal and infratidal environment represented by shallow-marine bioclastic wackestones locally named Champeh Member (Figure 10b). Figure 7. Position of the Fars Group in the schematically reconstructed Neogene Zagros foreland basin
before folding and uplift. Only distal parts of the basin are preserved in the current configuration of the ZFB as shown by the prograding sequences of the Fars Group, which correspond to the overfilled stage of the foreland basin. The proximal parts of the basin of the underfilled stage were eroded or underthrusted below the Sanandaj-Sirjan units of the upper Iranian plate.
Figure 8. (a) Isopachs of the Gashsaran, Mishan and Agha Jari Formations after a compilation by Motiei [1993]. (b) Thickness distribution of formations along eight profiles taken parallel to the transect shown in Figure 8a. The black solid line is calculated by averaging these eight profiles. Half arrows below the profiles aim at locating probable reverse faults that have influenced the position and importance of depocenters in the Zagros foreland basin during the Neogene. F denotes Firuz Abad.
These carbonates suggest the preservation of a narrow marine depression in the Central Part of the Zagros basin at this time.
[25] The thickness of the Gashsaran Formation decreases to 150 m close to the Persian Gulf (Figure 8). On the northern limb of the Zalemi anticline, the facies of the Gashsaran Formation is characterized by green-red shales, thin-bedded limestones deposited in intertidal to subtidal environments. At the Mountain Front, on the southern limb of the Asaluyeh anticline, the Gashsaran Formation is locally missing.
[26] The regional across-strike changes in facies reveal the development of a broad coastal plain, including tidal and deltaic deposits to the north, evolving southward to more marine environments. Across-strike thickness distribution further reveals a wedge-shaped subsidence profile typical of foreland basins opens toward the north, i.e., toward the tectonic load. The local abundance of carbonates indicates the presence of a narrow NW – SE-oriented subsident troughs bounded by tidal flats. Northward, it is bounded by a NW – SE-trending paleo-high at the current position of the Surmeh Fault. Variations in the thickness of the Gashsaran/Razak Formations across the Surmeh Fault sug-gest that deposition was controlled by the fault. The thinning, and locally the lack, of the Gashsaran Formation suggest forebulge-related uplift and/or basement thrusting at the Mountain Front.
4.2. Middle Miocene: Mishan Formation
[27] The Middle Miocene Mishan Formation is charac-terized by bioclastics and silty carbonates (Figure 10). This
sequence may be tentatively correlated with the eustatic rise in sea level reported at the Langhian-Serravalian boundary (16 Ma) (Figure 6). Close to Firuz Abad, the Mishan Formation consists of mudstones with bivalves and gastro-pods in association with patched reefal limestones charac-teristic of infratidal carbonates. The Mishan Formation coarsens upward and passes into bioclastic siltstones and sandstones alternating with shales indicating infratidal to intertidal environment (Figure 10). The thickness of the Mishan Formation increases from 600 m at the latitude of Firuz Abad to about 900 m southward (Figure 8). The thinnest part is found close to the Mountain Front with only 300 m. While the area south of the Karebass-Surmeh Fault is subsident, the northern Central Fars, limited by the Sabz-Pushan and the Karebass strike-slip faults, shows lower accumulation. Hence, by contrast with the Lower Miocene period, the northern Central Fars was uplifted over distance larger than 50 km supporting basement block faulting in the Middle Miocene.
[28] Structural pattern and facies variations in relation with the Meymand and Dalu anticlines deserve further consideration. Their bended axes outline the presence of an underlying active basement feature: the Sabz-Pushan right-lateral strike-slip fault [Berberian, 1995; Lacombe et al., 2006]. The Mishan Formation (Guri member) is repre-sented by reefal carbonates, which form linear ridges in the limbs of both folds (Figures 10c, 10d and 10e). Twenty kilometers northward of the Dalu anticline, these ridges disappear and the carbonates are replaced by more silici-clastics and thinner carbonates beds that are not distinguish-able from the Razak Formation. It is worth noting that the reefs disappear to the west of the Sim anticline, i.e., exactly above the trace of the NNW – SSE-trending Sabz-Pushan fault zone (Figure 10f). In addition, where Sim and Mey-mand fold axes connect, the reefal carbonates of the Mishan Formation (Guri Member) lie unconformably above the Gashsaran Formation (Champeh Member) (Figure 10g). We therefore infer that basement-involved deformation associated with transpressive right-lateral strike-slip faulting was active during the Middle Miocene. In summary, the depositional patterns and the facies variations within the Lower and Middle Miocene strata reveal wavelengths and amplitudes of tectonically-induced perturbations that could support basement-involved deformation.
5. Miocene-Pliocene Wedge-Top Basin:
Evidence for Inception of Cover Folding
[29] By the end of the Miocene, owing to continuing convergence, contraction resulted in shortening and folding of preorogenic margin sequences and syn-orogenic foreland sequences were deposited in synclinal depressions. The presence of intraformational unconformities within the Agha Jari Formation has been recognized for years in association with thin-skinned folding in the Dezful or Lorestan areas [e.g., Hessami et al., 2001; Sherkati et al., 2005]. However, no evidence for such unconformities has been described in the Central Fars so far. In addition, to our knowledge, no systematic structural study has been carried Figure 9. Sediment accumulation over the past 50 Ma
estimated from thickness distribution of synorogenic deposits shown in Figure 8. Upper and lower bounds refer to the northern and southern Fars basin, respectively. The black solid line is calculated by averaging both curves. It is clear that sedimentation drastically increase in the Neogene coevally with the collision initiation, shortening, erosion, and uplift in the internal part of the belt (see Figure 19).
Figure
1
0
out in the Central Fars to discuss the timing of folding. We aim at filling this gap by examining what are the uncon-formities that can be observed within the Fars Group and where they are positioned stratigraphically and across the ZFB. To this aim, we have performed numerous field campaigns from Shiraz to the Persian Gulf coast. Owing to the sediment filling, we had usually access to the limbs of anticlines only, lacking data in synclines and making the recognition of synfolding unconformities difficult.
5.1. Molasse Deposits: Agha Jari and Bakhtyari Formations
[30] The Agha Jari Formation usually displays an increasing upward abundance of red weathered shales and sandstones and the complete disappearance of carbonates. Isopachs of Figure 8 show that the Zagros foredeep was more subsident over this period than during the previous ones. The subsidence also increases significantly northward in agreement with the flexure of the Arabian margin undergoing tectonic loading.
[31] A NW – SE-trending high (300 m) and low (2100 m) across the Surmeh Fault suggests that basement-involved thrusting was ongoing in the Zagros foredeep at this time. This is also demonstrated in the bended syncline between the Dalu and Qarah folds. A Landsat image reveals the presence of an unconformity within the Agha Jari Forma-tion (Figure 10f). In the northern limb of the E – W-trending segment of the Dalu anticline, strata of the Agha Jari Formation onlap older strata in a direction oblique to fold axis. Because this angular unconformity is observed close to the trace of the Sabz-Pushan transverse fault we infer it may be related to the presence of a paleo-high in relation with the activity of this strike-slip basement fault at that time.
[32] In summary, the Agha Jari deposits were deposited in a flexural basin within which second-order basement-involved faulting occurred. In the studied area, the facies of the upper Agha Jari Formation laterally change from tidal red silts and shales, dominant to the south, to alluvial conglomerates in the north. In contrast, the Bakhtyari conglomerates, which have been deposited unconformably throughout the ZFB correspond to a period of relative inactivity in the foreland basin. The burial of the folded belt below the Bakhtyari conglomerates is consistent with an overfilled stage. Though the Agha Jari conglomerates
can be sometimes mistaken for the Bakhtyari conglomerates in the northern Fars, the mode of deposition of the Bakh-tyari conglomerates is different: the BakhBakh-tyari conglomer-ates blanket and fill a preexisting topography, for example, the synclinal valleys of folds or larger-scale depressions caused by basement-involved deformation. Such a type of deposition allows for the identification of the Bakhtyari Formation as it systematically onlaps toward the folds crest or directly lies unconformably.
5.2. Constraints on the Timing of Folding From Syn-Folding and Post-Folding Unconformities
[33] In the Lorestan province of the Zagros (Figure 1), where magnetostratigraphic constraints exist, the base of the Agha Jari Formation is dated as Serravalian-Tortonian (12 Ma) [Homke et al., 2004]. The ages of the upper Agha Jari Formation (including the Labhari Member), which is synfolding in the Lorestan, range between 7 and 3 Ma and the overlying conglomeratic layers of the Bath-tyari Formation, are younger than 3 Ma. Unfortunately, stratigraphic constraints on the continental molasse deposits are not yet available in our studied area. We have conse-quently used dating available in the Lorestan area. The maximum age uncertainty is expected in the northern part of the folded belt where continental deposits including Agha Jari and Bakhtyari Formations might be older as they were accumulated for a longer period of time in the more subsident part of the Zagros flexural basin. Taking into account the existing stratigraphic correlations made with reference to biostratigraphic constraints in the underlying carbonates of the Guri member (Figure 6) the continental formations are not older than 10 Ma.
[34] Additional constraints are provided by field obser-vations. The conglomerates, which are synfolding to the North of Shiraz (Figure 11a) are attributed to Plio-Pleisto-cene Bakhtyari Formation. These deposits are also overlain unconformably by younger conglomerates. If we consider that the Bakhtyari Formation is defined by a post-folding regional unconformity, the oldest conglomerates should belong to the upper Agha Jari Formation. In the absence of other time constraints, these observations provide lower and higher stratigraphic bounds for the synfolding strata of the upper Agha Jari Formation whose age has been set to 5.5 ± 2.5 Ma. The age of the base of the Bakhtyari
Figure 10. Facies and microfacies of the Fars Group in the northern part of the Zagros Folded Belt, in the vicinity of the Sim anticline (see Figure 2 for location). (a) Megaripples in red silts and sandstones of the Agha Jari Formation indicating tidal environment. Note that the older strata of the Razak Formation are very similar in this part of the ZFB. (b) Photograph of a thin section illustrating microfacies of shallow-water bioclastic carbonates belonging to the Champeh Member (Razak-Gashsaran Formation). Calcite twins in vesicles have been used to estimate differential stresses (Figure 4) across the whole ZFB [Lacombe et al., 2007]. (c) Structural relationships between Asmari, Razak and Mishan Formation on the northern of the Sim anticline (no evidence for unconformity has been found). (d) Western boundary of the reefal carbonates of the Guri member on the northern limb of the Sim anticline. (e) Massif corals in the Guri member. (f) Location map of the Sim anticline area near Meymand. (g) Structural relationships between the Guri reefal carbonates and the underlying carbonaceous member (Champeh member) of the Razak-Gashsaran Formation to the west of the Sabz-Pushan Fault Zone. The Guri member lies unconformably above the Champeh member where the Sim anticline connects with the Meymand anticline.
Figure 11. (a) Intraformational unconformities clearly depicted within the upper Agha Jari strata, on the southern limb of the Sorkh anticline (see Figure 2 for location). Above, the conglomeratic beds of the Bakhtyari Formation are onlapping the Agha Jari Formation. This suggests a slowing down of fold growth relatively to sedimentation rates. A subsequent stage of folding in association with the growth of the Qalat anticline is revealed by the tilting of the Bakhtyari conglomerates. (b) Intraformational unconformities within the upper Agha Jari strata on the southern limb of the Karbasi anticline (see Figure 2 for location). The cliff shows the upper alluvial conglomerates of the Bakhtyari Formation. They clearly onlap the older strata of the Agha Jari, again suggesting a slowing down of fold uplift. (c) Bakhtyari conglomerates overlying unconformably older strata of the Fars Group and carbonates of the Asmari Formation on the southern limb of the Khunj anticline indicating significant erosion prior to the deposition of the conglomerates. (d) Folded Bakhtyari conglomerates on the southern limb of the Takhteh anticline.
Formation is considered to be 3 ± 0.5 Ma throughout the belt. The age of the top of the Bakhatyari conglomerates has been roughly approximated by taking advantage of the relation between thickness and the extrapolation of maxi-mum sedimentation rates (Figure 9). The absolute ages presented in the following sections should be considered as indicative estimates for forthcoming studies on fold kinematics.
5.2.1. Sorkh and Qalat Anticlines: Northern Fars [35] Twenty kilometers to the North of Shiraz and just south of the High Zagros Fault (Figures 3 and 11a), above marls, sands and carbonates of the Mishan Formation, conglomerates attributed to the Upper Agha Jari show progressive unconformities revealing syn-sedimentary rota-tion on the southern limb of the Sorkh anticline. The upper part of the outcrop shows conglomerates that are dipping toward the north on the northern limb of the Qalat anticline. These latter are consequently onlapping older conglomer-ates and would better correspond to the younger Bakhtyari Formation. We deduce that the last increment of deforma-tion of the Qalat anticline mainly postdates the development of the Sorkh anticline, which initially grew by limb rotation during the deposition of the upper Agha Jari conglomerates at 5.5 ± 2.5 Ma. This episode is followed by a decrease in the fold uplift rate relative to the rate of accumulation of upper Bakhtyari conglomerates. To figure out the age of the Qalat fold we extrapolate the thickness of strata to the average sedimentation rates (Figure 8). Taking into account that at least 100 – 200 m were deposited, we obtain a maximum age of 2.5 2 ± 0.5 Ma for the top of the conglomerates and hence for the last folding event. 5.2.2. Karbasi Anticline: Central Fars
[36] Progressive unconformities are also found in the southern limb of the Karbasi anticline about 200 km southeastward of the Sorkh anticline (Figures 3 and 11b). The progressive unconformities are found in the lower and older strata belonging to the upper Agha Jari Formation. The cliff formed by the Bakhtyari conglomerates again displays no evidence for intraformational unconformity. We thus conclude to a similar Pliocene sequence of folding as for the Sorkh anticline. After Karbasi anticline developed by limb rotation 5.5 ± 2.5 Ma ago, the Bakhtyari conglom-erates were deposited unconformably during a period of relative folding quiescence.
5.2.3. Khunj Anticline: Southern Fars
[37] The Khunj anticline, in the Central Fars, is located in an elongated faulted belt aligned along the Surmeh fault (Figures 3 and 11c). According to existing geological maps and satellite images its southern limb is suspected to be faulted. However, we observe a remnant patch of Bakhtyari conglomerates that overlie unconformably the folded strata of the Gashsaran and Asmari formations. Although no structural relationship has been found between the Agha Jari Formation and the Bakhtyari conglomerates, we suspect
that the latter lie unconformably above the Agha Jari Formation. The Bakhtyari conglomerates are made of pebbles of carbonates, which are essentially flat. In contrast with the Qalat anticline, but similarly with the Karbasi anticline, the main phase of folding seems to have ceased before the deposition of the Bakhtyari Formation at 2.5 2 ± 0.5 Ma.
6. Constraints on Quaternary Fold Growth
and Its Relationships With Active Basement
Faulting: Insights From Morphology, Drainage
Pattern, and Historical Seismicity
6.1. Regional Versus Local Drainage Pattern: The Role of Thick-Skinned Deformation
[38] Studies dealing with morphological patterns in the Zagros Folded Belt date back to tens of years with reference to the drainage pattern observed in the Dezful area [Oberlander, 1985; Tucker and Slingerland, 1996]. The main transverse streams observed in the Zagros Folded Belt form spectacular river gorges in the Dezful area [Oberlander, 1985]. Then they join the Tigris-Euphrates rivers or flow directly into the Persian Gulf. Remarkably, the Central Fars is characterized by a lack of evidence for river antecedence (or superposition) and N – S transverse drainage throughout the folded belt (Figure 12). Rivers do not follow a simple south-flowing course across the folded belt and have a course shorter than axial rivers. This drainage pattern suggests that rivers are strongly controlled by the folds growth. Deviation of the course of rivers may be explained as the result of transverse rivers being dammed by the uplift of major anticlines.
[39] It should be noted that the sinusoidal shape of folds is superimposed on background uplift controlled by base-ment involvebase-ment that is outlined by a regional topography sloping at 0.3° – 0.5° toward the foreland [Mouthereau et al., 2006]. The drainage network is thus expected to be con-trolled to some extent by thick-skinned deformation. For instance, north of the Karebass-Surmeh faults segment, most rivers are forced to flow axially toward the east. This is the response of the tilting of the regional base level toward the east caused by the deformation associated with a set of oblique transpressive ridges: the Surmeh-Karebass and the Sabz-Pushan faults segments (Figure 12).
[40] In the center of the Fars arc, east of longitude 53.5°E, we identify a domain where the drainage patterns do not show any preferential flowing orientation and where most of the northern transverse rivers rapidly die out in several salt depressions. This endoreic fluvial system is discon-nected from the main western fluvial system debouching in the Persian Gulf. Such intermontane depressions, which are associated with a decrease of channel gradient, reopens
Figure 12. River drainage network in the Zagros Folded Belt reconstructed on the basis of Landsat images. Inset in the lower left show drainage network in the absence of structural information. The rivers flowing toward the south are closely related with largest slope gradient near active basement-involved fault zones. Away from these domains, rivers often flow axially or die out into the Central Fars depression.
progressively to the south as the slope, increasing in the footwall of the active Surmeh fault, allows rivers to flow again transversally. Again southward, the rivers flow axially to the West, and merge before debouching in the Persian Gulf, north of the Mountain Front Fault.
[41] As already inferred from observations of isopachs (Figure 8), we propose Quaternary basement-involved shortening produced topographic highs and lows that con-trolled the regional course of rivers. Growing folds likely secondarily created abnormal southeast or southwest flow-ing streams by forcflow-ing rivers to flow axially. This is particularly clear in the vicinity of the Dalu anticline to the northern Central Fars and the fold system of Hava,
Zalemi or Asaluyeh anticline closer to the Mountain Front (Figure 12).
6.2. Fold Uplift and Basement Faulting: Geomorphic and Seismotectonic Constraints
[42] The degree of removal of weakly cemented and siliciclastics strata of the Fars group above the competent beds of the Asmari Formation may serve as a proxy for relative age of folds. To a first approximation, assuming initially constant cover thickness, fold growth rates and river stream power, the oldest anticlines would be those where Mesozoic rocks crop out owing to more bedrock uplift or incision (Figure 13). However, several other Figure 13. Digital elevation model of the Fars Province (northern domain) based on SRTM data with
90 90 m resolution. Anticlines mentioned in this study are shown together with the location of axial and transverse topographic profiles. Major strike-slip fault zones that affect the sedimentary cover are also depicted. Black dashed ellipsoids correspond to the areas destroyed during major earthquakes of years 1824 near Qalat and 1972 (Qir earthquake) after Ambraseys and Melville [1982].
parameters may additionally control the geomorphic signa-ture of folds in foreland such as (1) the ratio of the sediment accumulation to the structural uplift, (2) the balance of the stream power versus the bedrock resistance and the rates of fold uplift [Burbank et al., 1996; Tucker and Slingerland, 1996] or (3) the initial thickness of weakly cemented rocks. [43] We have noticed that one remarkable feature of the Central Fars is the lack of rivers cutting through the anticlines, with noticeable exceptions in areas where base-ment uplifts have been recognized. This would imply that the structural uplift was too rapid for rivers to cut down through the fold. Moreover, a simple in-sequence cover folding development in the cover is unlikely in the Central Fars as it would have certainly resulted in antecedence or superposition of rivers through the folded belt as exempli-fied in the Apennines [Alvarez, 1999]. This would further imply that all folds possibly formed nearly at the same time. Hereafter, we focus on this latter possibility and the possible control by inherited fault-controlled basement highs. 6.2.1. Qalat Anticline
[44] The Qalat anticline is one of the few folds for which a significant part of contraction seems to postdate the deposition of the Bakhtyari conglomerates. It developed in the southern limb of the Sorkh fold revealing, at least locally, a southward in-sequence propagation of folding. The fold wavelength, defined by the distance between two adjacent synclinal axes, is 13 km consistent with the typical fold wavelengths in the Central Fars (Figure 5). A NW – SE axial topographic profile shows that the minimum total uplift of the Asmari surface is 1300 m relative to the eastern local base level. Along the axial profile we also notice a differential topographic elevation of local base levels: the eastern nose of the fold is 400 m lower than the western one (Figure 14) and the latter shows a wind gap within the Asmari carbonates. This illustrates the ongoing westward propagation of folding. Since the abandonment of the river, the fold was uplifted by about 200 m.
[45] Given a minimum structural uplift of the Asmari surface of 1300 m relatively to the eastern local base level
and an age of 2.5 2 ± 0.5 Ma for the inception of fold uplift, we deduce that fold uplifted at a rate of 0.58 ± 0.22 mm/yr. In addition to thin-skinned folding, we chal-lenge that the morphology of the Qalat anticline might have been also influenced by basement-involved deformation. Indeed, a destructive earthquake (Ms= 6.4) that occurred in 1824 (25 June) damaged the city of Shiraz and the city of Qalat was ruined and a few hundred people were killed [Ambraseys and Melville, 1982]. Berberian [1995] noted that the destruction area related to the earthquake is elon-gated NNW – SSE, parallel to the northern extension of the Sabz-Pushan transverse fault. We infer that the current morphology of the Qalat anticline might be largely related to vertical movements along the right-lateral Sabz-Pushan fault zone.
6.2.2. Qarah and Dalu Anticlines
[46] The Qarah and Dalu anticlines are subparallel and distant by 15 – 25 km (Figures 10 and 13). The axial topographic profile across the Qarah anticline shows important denudation in the center of the anticline. Its wavelength is abnormally larger than 20 km. A comparison of the transverse topographic profiles 1, 2, 3 with the axial profile provides an estimate for the volume of pre-Asmari rocks removed by erosion, which is of the order of 60 km3 (Figure 15). The Mesozoic strata are exposed within the anticline in two large wind gaps, which are weakly elevated in regard to adjacent base levels. This would suggest the absence of recent fold uplift. The Dalu fold, to the south, is a drag fold that developed above the active Sabz-Pushan right-lateral strike-slip fault zone (Figures 10 and 13). By contrast with the Qarah anticline, the Dalu anticline shows much less denudation as illustrated by the slightly dissected structural surface of the Asmari carbonates (Figure 15). Its fold wavelength of 12 – 15 km is also shorter and the minimum uplift of the Asmari surface with respect to the local base level is larger, i.e., 1800 m on average. The presence of a water gap in the eastern nose of the fold suggests that eastward propagation of the fold is still ongoing, away from the Sabz-Pushan fault zone. These Figure 14. Axial topographic profile across the Qalat anticline. See Figure 13 for location.
Figure 15. (left) Transverse and (right) axial topographic profiles of the Qarah, Dalu, Takhteh, and Shanul folds. See location in Figures 13 and 16.
geomorphic data may suggest that the Dalu fold may be younger than the Qarah fold.
[47] In 1892, a macroseismic earthquake have been reported exactly in this bended portion of the Dalu anticline [Berberian, 1995], which confirms the still ongoing defor-mation along the Sabz-Pushan basement fault. Because isopach maps and field observations both indicate that this area was controlled by the activity of the Sabz-Pushan fault zone as soon as the Middle-Upper Miocene times, we suggest that the sequence of thin-skinned folding between Qarah and Dalu anticline was also influenced by thick-skinned deformation along the NNW – SSE-trending trans-pressive ridge of the Sabz-Pushan Fault Zone. This may explain (1) the abandonment of transverse rivers cutting through the Qarah anticline and (2) to some extent, the origin of the Dalu anticline itself.
6.2.3. Takhteh Anticline
[48] The Takhteh anticline lies within a girdle of en-e´chelon folds aligned along the trend of the active Surmeh fault (Figures 11d, 14, and 15). It is characterized by a fold wavelength of 13 14 km and the presence of folded Bakhtyari conglomerates. The current activity of the Sur-meh fault is revealed by the 1972 Qir earthquake (Ms= 6.9), which killed about 5000 people [Ambraseys and Melville, 1982]. Similarly with the Qarah anticline, the Mesozoic cover is well exposed within the Takhteh anticline indicat-ing that the fold was deeply eroded. Differential N – S elevation across the fold (Figure 15) is explained by the axial west-flowing river, which has wiped out the Bakhtyari conglomerates deposited in the northern fold limb. This contrasts with the southern limb where the Bakhtyari con-glomerates are thick and form cliffs (Figures 11d and 15). A significant part of the shortening postdates the deposition of the Bakhtyari conglomerates such as for the Qalat fold (Figure 11d). To explain the apparent inconsistency between the smooth fold morphology and the fact that Bakhtyari Formation is folded, it is important to consider that the Takhteh anticline developed in the vicinity of a Miocene paleo-ridge (Figure 8). The Cenozoic strata were initially thinner than in adjacent areas so it may have been more rapidly incised by rivers. On the basis of extrapolation of sedimentation rates, and taking into account that the pre-folding Bakhtyari conglomerates are not thicker than 200 m, we deduce that the top of the conglomerates should be2 ± 0.5 Ma. Given a minimum total uplift of 700 m, we infer a maximum uplift rate of 0.33 ± 0.25 mm/yr.
6.2.4. Shanul Anticline
[49] The Shanul anticline is located 40 km to the south of the Takhteh anticline (Figures 15 and 16). It has a wave-length of 13 – 14 km comparable with Qalat, Dalu and Takhteh folds. A major difference however is that its structural topography is outlined by the resistant carbonates of the Mishan Formation. Transverse and axial topographic profiles show that the Mishan Formation is not significantly eroded (Figure 15). Given the altitude of the Mishan carbonates, the fold uplift with respect to the adjacent local base levels is about 900 m. To explain this particular morphologic pattern we must again refer to the isopach maps (Figure 8). They show that this area was a main
Mishan depocenter in the footwall of the Surmeh Fault. As the rivers incised a thicker Neogene sequence, only a limited part of the upper layers overlying the competent Mishan carbonates were removed. This is consistent with the fact that in contrast with folds located in the hangingwall of the Surmeh Fault, the local base levels at both noses of the fold reveal little incision by rivers and even deposition in the large synclinal valleys. We suggest that the age of folding may not be significantly different from that of Qalat, Dalu and Takhteh anticlines.
6.2.5. Mountain Front
[50] The deformation at the Mountain Front close to the Persian Gulf is characterized by a 200-km-long girdle of five en-echelon anticlines in the hangingwall of the Moun-tain Front Fault (Figure 16). The current activity of this major basement reverse fault is revealed by the series of recent historical and instrumental earthquakes: two macro-seismic earthquakes have been reported in 1883 (Ms= 5.8) whose destruction area is aligned in the direction of the Asaluyeh anticline (Profile 1) and more recently in 1950 (Ms = 5.5) [Ambraseys and Melville, 1982; Berberian, 1995].
[51] The anticlines have wavelengths of about 20 km, on average, larger than Qalat, Dalu, Takhteh and Shanul anti-clines but close to the wavelength of the Qarah anticline. The maximum total uplift of the folds with respect to the Persian Gulf increases from 800 m (profile 5) and 1400 m (profile 1) northwestward (Figure 17). The remarkable differential elevation of 200 to 700 m of local base levels across the Mountain Front folds is controlled to first-order by basement thrusting along the Mountain Front Fault. Furthermore, local tilting of Quaternary terraces near the Mountain Front [e.g., Mann and Vita-Finzi, 1988], geodetic surveys [Walpersdorf et al., 2006] as well as14C dating of Holocene fluvio-marine terraces [Oveisi et al., 2007] shows that a significant part of the deformation in the Zagros is currently accommodated across the Mountain Front.
7. Discussion
7.1. Thin-Skinned Buckling of the Sedimentary Cover [52] The present study has emphasized particular features of folding: (1) the existence of characteristic fold wave-lengths, (2) no thrust is associated with folding at the surface, (3) across the 200-km-width ZFB, folding devel-oped rather synchronously over the past 5.5 ± 2.5 Ma, and (4) the progressive onlaps of conglomerates onto older growth strata (Figure 11) suggest a slowing down of fold growth relatively to the rate of deposition. To explain such results and observations we challenge that the development of the Zagros fold train is better reproduced by invoking buckling of the cover, while both distributed and localized thrusting occurred within the basement.
[53] The theory of buckling predicts that a single com-petent layer with random perturbations overlying a weaker matrix will develop into a regular fold train when subjected to layer-parallel shortening [Biot, 1961; Zhang et al., 1996]. Fold train results from the process of selection and ampli-fication of initial perturbations that is dependent on the
competency contrasts between matrix and layer. The initi-ation of fold growth is exponential when the dominant wavelength is selected and then gradually slows down.
[54] Several analytical and numerical models have con-sidered folding in the Zagros fold belt as a case example for studying mechanisms of buckling [Biot, 1961; Schmalholz et al., 2002; Turcotte and Schubert, 2002]. The thick Paleozoic and Mesozoic carbonates can be treated as a single layer overlying a homogeneous and finite viscous matrix lying in the Cambrian salts. It has been noticed for years that the pure elastic solution for buckling wavelengths given by 2pHpffiffiffiffiffiffiffiffiffiE=s often requires unrealistic high layer-parallel stress s to fit observed buckling wavelengths. However, folding wavelengths may be roughly approxi-mated by a pure elastic solution [Mouthereau et al., 2006] assuming a layer-parallel stresss of at least 300 MPa and a
low value of Young’s modulus E of 109Pa supported by recent mechanical experiments on Asmari carbonates [Amrouch et al., 2005]. The result is however largely dependent on the effective elastic thickness H adopted for the competent layer. A value of less than 2 km, i.e., representing only a quarter of the value of the unbended layer, is compatible with fold curvatures (>10 5m 1) and brittle yielding in a plastic(brittle)/elastic layer. On the other hand, a pure viscous solution for buckling wavelengths given by 2pH6 1/33pffiffiffiffiffiffiffiffiffiffiffiffihl=hmis limited by viscosity contrasts between layer and matrix of less than 2 orders of magni-tudes [Schmalholz et al., 2002]. The viscosity of the Hormuz evaporites has been fixed to typical viscosities (Newtonian) of salts between 1017 and 1018 Pa s [Mouthereau et al., 2006]. In this case, fold wavelengths may be reproduced using a thin viscous layer H of 2 km and Figure 16. Digital elevation model of the Fars Province (southern domain) based on SRTM data with
90 90 m resolution. Anticlines mentioned in this study are shown together with the location of axial and transverse topographic profiles (Figure 15). Black dashed ellipsoids correspond to the areas destroyed during major seismic events that occurred in 1972 (Qir earthquake) and 1950 after Ambraseys and Melville [1982].
layer viscosity hl of 1019 Pa s. We see that none of these end-member models are satisfactory since they consider that only a thin portion of the total thickness of the sedimentary cover is competent. A viscoelastic solution, which consists of an elastic sublayer at shallow depths overlying a viscous (non-Newtonian) sublayer with finite thickness to greater depths seems to reproduce better fold wavelengths [Schmalholz et al., 2002].
[55] In details, the wavelengths of folds are not perfectly homogenous. On the basis of cross section (Figure 4), wavelengths vary from 11 km in the vicinity of the Surmeh fault to a maximum of 30 km in the High Zagros. To the south, toward the MFF, fold wavelengths reduces again to 17 km on average. It is not surprising that the observed wavelengths may be lower than those predicted by models of buckling. After inception of buckle folding and selection of a dominant wavelength, shortening increases thus reduc-ing the wavelengths with time as fold amplitudes increase. We infer that folding initially developed with higher wave-length/amplitude ratio across the ZFB. Then as shortening increased and wavelengths reduced, limbs rotated and amplitudes increased. For instance, the ratios of the wave-length to folds amplitude range from 15 (30:2) in the IZ whereas it decreases to 5.5 (11:2) when approaching the Surmeh Fault (Figure 4). In the IZ, the folds have preserved their initial geometry whereas in the ZFB more shortening led to reduce the wavelengths.
[56] Even though the kinematics of folding cannot be accurately and continuously tracked over time owing to the lack of sufficiently accurate stratigraphic data, there are enough constraints for deciphering the initiation of folding associated with syn-folding sedimentation and limb rotation at 5.5 Ma as well as the subsequent growth stage of folding. One can draw the following scenario (Figure 18). In the Upper Miocene – Early Pliocene, after 5.5 ± 2.5 Ma, folds
initially developed rapidly as buckle instabilities with wave-lengths larger than 25 km and negligible amplitudes (Figure 18a). As shortening increased, fold wavelengths progressively reduced. Folds grew by limb rotation in association with intraformational unconformities within the upper Agha Jari Formation (Figure 18b). This stage is similar to typical detachment folding and models of folding with constant limb length [Storti and McClay, 1995; Ford et al., 1996]. As fold uplift rate decreased relatively to the rate of sedimentation, Bakhtyari conglomerates were deposited unconformably (Qalat, Karbasi, Khunj anticlines) (Figure 18b). By 3 Ma, some folds (Qalat, Dalu, Takhteh, Shanul, Mountain Front) rejuvenated with fold uplift rates of the order of 0.3 – 0.6 mm/yr. (Figure 18c) This last event is observed in portions of the ZFB where underlying basement deformation occurred and is still ongoing and may thus explain, at least partially, folding. Alternatively, the erosion on the crest of anticlines driven by climatic forcing in the past 3 Ma may have also enhanced folding. 7.2. Sequence of Thin/Thick-Skinned Deformation in the Zagros: From Paleocene Foreland Deposition to Cover Buckling at the Arabia/Eurasia Plate Boundary
[57] Recent studies have pointed out that the outward propagation of the ZFB may have occurred either in a single thin-skinned mode or in a two-step evolution involving a period of thin-skinned deformation followed by a period of thick-skinned deformation. The present study suggests that the sequence of deformation in the whole crust was prob-ably more complex.
[58] The first model was proposed by Hessami et al. [2001] who suggested, on the basis of the interpretation of local angular unconformities, that folding in the Central Fars started at the end of the Eocene and then propagated with typical thin-skinned forward-propagating sequence Figure 17. Transverse topographic profiles across-folds associated with the Mountain Front. See
location in Figure 16.
southward. Then Molinaro et al. [2005] suggested that the ZFB resulted from two distinct periods: the first being characterized by in-sequence thin-skinned mode since the Middle-Late Miocene and the second by out-of-sequence thick-skinned thrusting since the Pliocene. However, one can draw several limitations to both interpretations. First, the occurrence of in-sequence propagation is not so common in folded belts controlled by a ductile de´collement
as they often shows rearward propagation [Costa and Vendeville, 2002], oscillating sequence of deformation [Smit et al., 2003] or very fast propagation of the wedge front [Bahroudi and Koyi, 2003]. Second, none of the available studies satisfactorily account for the coeval involvement of cover (thin-skinned) folding and basement (thick-skinned) deformation currently observed.
7.2.1. From Obduction to Collision
[59] In Figure 19, we synthesize the sequence of defor-mation from the obduction to the inception of the Zagros collision. The sub-marine obduction of the ophiolites initi-ated in the Turonian [Ziegler, 2001] and lasted until the upper Cretaceous (Campanian). Since then, a narrow trough, filled with up to 1 – 2 km of hemipelagic carbonates and siltstones of the Gurpi Formation (Maastrichtian), developed to the south of the present position of the MZT [Koop and Stoneley, 1982]. This basin reveals the inception of flexural subsidence resulted from the subaerial and ongoing outward advancement of ophiolitic thrust nappes onto the Arabian continental margin. Following this epi-sode, from the Upper Cretaceous—Paleocene and until the Eocene, the environments of deposition significantly changed and the subsidence increased [Koop and Stoneley, 1982]. The deposition of 1-km-thick siliciclastic turbidites (Amiran Formation) in the Lorestan province suggests ongoing propagation of the ophiolitic nappes southward [Homke et al., 2007]. Onto the Fars promontory, the turbiditic sequences pass laterally into continental weath-ered shales or shallow-marine clastics of the Pabdeh-Sachun formations. This region represents the distal, nearly unde-flected, portion of the foreland basin. Foreland deposits are overlain by Eocene reefs and dolomites of the Jahrom Formation in the Fars platform passing to Eocene turbiditic sequences as reported on top of the obducted units north-ward [Hempton, 1987; Beydoun et al., 1992] suggesting a flexurally controlled subsidence.
7.2.2. Inception of Collision and Development of the Thin-Skinned/Thick-Skinned Collision Belt
[60] The early Neogene period is characterized by the deposition of carbonates throughout the Arabian continental margin (Asmari Formation) and the Iranian plateau (Qom Formation). These are intertidal boundstones, micritic and foraminiferal limestones, including mainly nummulites and miliolids within the Asmari Formation. Biostratigraphic constraints yield ages ranging from Upper Oligocene to Lower Miocene for both formations (Chattian-Aquitanian boundary) [Schuster and Wielandt, 1999]. The Iranian Plateau and the Zagros Folded Belt, which were below sea level at that time, are separated from each other by a NW-Figure 18. Schematic model of fold development from (a)
buckling stage to (b, c) detachment folding involving limb rotation. The gray layer is a schematic representation of the competent (viscous-elastic-plastic) folded cover rocks. The Agha Jari and Bakhtyari Formations are shown as thin white layers overlain by a thick layer with black dots, respectively. R is the ratio of the wavelength to folds amplitude, andl is the wavelength of folds.
Figure 19. Reconstruction of the Zagros evolution from obduction to collision. Mesozoic strata, are given in green color, Paleogene in orange and the Neogene in yellow (see inset captions for details). White arrows pointing upward indicate uplift whereas black arrows pointing downward are for subsidence. Erosion and orientation of sediment supply from uplifted areas is indicated by smaller black arrows. ZFB, Zagros Folded Belt; IZ, Imbricate Zone. Note that since 3 – 2 Ma the Imbricate Zone and the MZT are inactive. The current seismogenic deformation has been transferred southward toward the Mountain Front, which attests for southward propagation of the basement wedge and part of the convergence accommodation across the Mountain Front.
Figure 19
![Figure 2. Geological map of the Fars based on a compilation of geological 1:250,000 and 1:100,000 scale maps from the National Iranian Oil Company [1977]](https://thumb-eu.123doks.com/thumbv2/123doknet/14794133.602839/4.918.140.770.110.636/figure-geological-fars-compilation-geological-national-iranian-company.webp)
![Figure 8. (a) Isopachs of the Gashsaran, Mishan and Agha Jari Formations after a compilation by Motiei [1993]](https://thumb-eu.123doks.com/thumbv2/123doknet/14794133.602839/11.918.158.769.107.990/figure-isopachs-gashsaran-mishan-agha-formations-compilation-motiei.webp)
